As efforts to screen women using mammography have increased, so has the number of women diagnosed with breast carcinoma in-situ (BCIS). Clarification of the role of mammographic characteristics, such as mammographic breast density (MBD), with BCIS risk is important as 1) Up to 20% of screened breast cancer patients are diagnosed with this lesion and 2) increased MBD is associated with a 2-6 fold increase in the risk of invasive breast cancer. This project will study this association for the first time by collecting and reviewing mammograms in the largest population-based, case/control study of BCIS to date. The sample includes all cases of female BCIS aged 20 to 79 years diagnosed among residents of Connecticut from 9/15/94-3/14/98 as well as a series of age-matched random-digit-dial controls; these women (875 ductal carcinoma in-situ cases, 123 Iobular in-situ cases and 999 controls) are followed as part of our BCIS outcomes study. Information already collected includes 1) questionnaire data on family history of cancer, pregnancy menstrual history, exogenous hormone use, as well as cancer screening history, 2) follow-up interview data on lifestyle and medical outcomes, and, for cases, 3) paraffin embedded tumor blocks for immunohistochemical testing of ER, PR, p53 and HER-2/neu receptors, as well as 4) blood-samples for testing of mutations in BRCA1/BRCA2. We will examine whether 1) increased MBD or change in MBD is associated with BCIS risk, and 2) whether any correlation exists between clinical variables, mammographic characteristics, and outcome. This information will be used to 3) refine our models of breast cancer risk and to 4) examine whether mammograms can be characterized with respect to high and low risk outcomes. The addition of mammograms to a pre-existing epidemiologic cohort notable for its unique subject compositition, high participation rates, centralized pathology review and multi-disciplinary data collection, is a cost-efficient and scientifically valuable opportunity.